Andenov/proact

AI-powered anticipatory action platform for climate risk and farmer alerting in Uganda # PROACT — Anticipatory Action Platform An AI-powered climate risk and farmer alerting system for Uganda. ## What It Does - Monitors 5 pilot districts in Northern and Eastern Uganda - Computes daily flood, landslide, and food stress risk scores using Open-Meteo weather data - Displays district-level risk on an interactive dashboard with map, alerts, and trend charts - Sends SMS alerts to enrolled smallholder farmers (mock provider by default) ## Quick Start ### Prerequisites - Docker + Docker Compose - OR: Python 3.11+, Node 20+, PostgreSQL 15 with PostGIS ### Option A — Docker (recommended) ```bash cp .env.example .env docker-compose up --build ``` Then visit: - Dashboard: localhost - API docs: localhost ### Option B — Local development **1. Database** ```bash # Start PostgreSQL with PostGIS locally, then: createdb proact ``` **2. Backend** ```bash cd apps/api python -m venv .venv source .venv/bin/activate # Windows: .venv\Scripts\activate pip install -r requirements.txt cp ../../.env.example .env # Edit .env with your DATABASE_URL uvicorn app.main:app --reload --port 8000 ``` On first start, the API automatically: - Creates all tables - Seeds 5 pilot districts and hazard features - Seeds 12 sample farmers - Creates admin user: `admin@proact.org` / `proact2024` **3. Frontend** ```bash cd apps/web npm install NEXT_PUBLIC_API_URL=localhost npm run dev ``` Visit localhost ## Running the Data Pipeline After starting the API, trigger data ingestion: ```bash # Fetch weather data from Open-Meteo (last 37 days + 7-day forecast) curl -X POST localhost # Compute risk scores for all districts curl -X POST localhost # Generate alerts from risk scores curl -X POST localhost ``` The scheduler also runs this pipeline automatically at 06:00 UTC daily. ## Project Structure ``` proact/ apps/ api/ FastAPI backend …